Role of age at onset in the clinical presentation of bipolar disorder in Indian population.

OBJECTIVE: The objective of this study was to determine any association of age at onset (AAO) with clinical presentation of bipolar disorder (BD) and family history of illness.
MATERIALS AND METHODS: A hospital-based cross-sectional observational study was conducted including 162 patients having a diagnosis of BD current episode manic. Individuals were divided into three subgroups according to AAO, i.e., early-onset BD (EOBD) (AAO ≤21 years), intermediate-onset BD (AAO - 22-34 years), and late-onset BD (AAO ≥35 years). The subgroups were compared on clinical variables; items of the Young Mania Rating Scale (YMRS), Hamilton Depression Rating Scale (HAM-D), and Scale for the Assessment of Positive Symptoms (SAPS); and family history of illness.
RESULTS: The early-onset group had significantly more episodes per year than the other groups (P < 0.001). The prevalence of family history of mood disorder was also significantly higher in the early-onset group than the other subgroups. AAO was found to be significantly associated with different items of YMRS, HAM-D, and SAPS. The early-onset group had higher rating on irritability, motor activity-energy, sexual interest, depressed mood, delusions, and thought disorders, whereas the late-onset group had higher rating on elevated mood.
CONCLUSION: EOBD can be considered as a specific phenotype of BD, which is more homogenous, severe, and inheritable form of illness. Copyright:

Bipolar disorder (BD) is a chronic mental disorder characterized by alternating episodes of depression, euthymia, (hypo) mania, and mixed states.[1] The usual age of onset of BD is before 30 years, and almost 90% of patients have onset before 50 years of age.[23] BD has been reported to have a genetic basis with heritability of at least 60%–85%.[45] A number of studies have attempted to identify the susceptibility loci adding evidence of linkage in multiple genomic regions.[6] However, these genetic studies have given conflicting results; which may be due to the lack of consensus regarding the proper definition of the affected phenotype and the questionable homogeneity of bipolar illness.[456] Thus, attention is shifting toward the study of clinical indicators of BD that are familial and may be useful for identifying inheritable forms of the illness.

Classifications such as the Diagnostic and Statistical Manual of Mental Disorders-5 defined certain specifiers of BD that have proven phenomenological relevance.[7] However, most are of limited utility in identifying homogeneous subgroups. There is a necessity to identify reliably defined clinical features that are clinically valid and associated with underlying genetic process. These features can help in identifying genotypes and making informed treatment decisions.[8]

There is considerable evidence to suggest that the clinical expression of BD differs according to age at onset (AAO), which has therefore been identified as a potential specifier of interest, especially patients having early onset of illness have specific clinical features and outcomes different from those of later onset of illness.[8910] In this regard, multiple admixture studies have revealed that such patients could be categorized into three homogeneous subgroups (early-onset BD [EOBD], intermediate-onset BD [IOBD], and late-onset BD [LOBD]).[1112131415161718]

By making subgroups based on these cutoff points, the present study was aimed to compare the clinical presentation using the items of the Young Mania Rating Scale (YMRS), Hamilton Depression Rating Scale (HAM-D), and Scale for the Assessment of Positive Symptoms (SAPS) and to compare the presence of family history of the illness in Indian population. Such subgrouping may help in identifying phenotypes among BD patients.

MATERIALS AND METHODS

A hospital-based cross-sectional observational study was conducted at the psychiatry department of the tertiary care hospital of western India. The study was conducted over a period extending from June 2018 to December 2018. Ethical clearance was taken from the research review board and ethical committee of the institution. Informed written consent was obtained from all the legal guardians of participants before inclusion in the study.

A consecutive sample of 162 patients admitted to the inpatient department with a diagnosis of BD current episode manic (as per the Diagnostic and Statistical Manual of Mental Disorders-5 criterion) was included in the study. The diagnosis was confirmed by applying the Mini-International Neuropsychiatric Interview (MINI).[19] Patients with history of intake of any psychiatric medicine in the last 2 weeks were excluded from the study because it can alter the severity of the symptoms. Patients suffering from mental retardation, dementia, comorbid psychiatric and substance-related disorder (as confirmed by using the MINI), and any other acute medical illness were excluded from the study.

Study participants were divided into three subgroups according to AAO, namely EOBD (AAO ≤21 years), IOBD (AAO = 22–34 years), and LOBD (AAO ≥35 years). There were 67, 59, and 36 patients in early, intermediate, and late-onset groups, respectively. The cutoff for AAO was determined on the basis of a review article by Geoffroy et al.[20] The cutoff value of 21 years for early onset was also recommended in other studies.[2122]

AAO was defined as “the age at which the patient first met criteria for a depressive episode, mania or hypomania as indicated by patient history, recall of family members and medical records.” This definition is close to the exact beginning of the illness and had been found to be highly reliable.[23]

After applying stringent inclusion and exclusion criteria, participants were subjected to the tools of study. Clinical interviews and symptom ratings according to the YMRS,[24] HAM-D,[25] and SAPS[26] were performed at the first presentation in the outpatient department.

These subgroups were compared with the following tools of the study:

Sociodemographic and clinical variables – A semi-structured pro forma was used to record the sociodemographic variables including age, sex, residence, occupation, marital status, monthly income, and education. Items occupation, education, and monthly income were adopted from the modified Kuppuswamy's socioeconomic scale.[27] This information was gathered from first-degree relatives and the patient himself

Clinical assessment – The clinical variables were gathered through a clinical interview of participants and their immediate family members. The selection of clinical variables was based on literature review[28] and clinical observation. The clinical variables included number and duration of the illness (chronicity) and AAO of the illness

Family history of mood disorder in first-degree relatives – Familial psychiatric illness was investigated using the Family Interview for Genetics Studies[29] applied by a trained psychiatrist. A complete family history of psychiatric illness of first-degree relatives was obtained from each proband and at least one first-degree relative. The information was supplemented by medical reports of relatives.

Symptom-related variables – It was assessed using YMRS, HAM-D, and SAPS. Items were removed from the rating scales that were a priori considered redundant among them (i.e., similar symptoms rated more than once). Like, the HAM-D scale ratings of early, late, and middle insomnia; psychomotor agitation; and impaired insight were removed in favor of the YMRS ratings of items decreased sleep, disruptive or aggressive behavior, and impaired insight, respectively. The SAPS behavioral rating domain was removed in favor of YMRS ratings of disruptive or aggressive behavior and inappropriate appearance. The YMRS rating of content and language-thought disorder was removed in favor of the more comprehensive rating of delusion, hallucination, and thought disorder on the SAPS. Thus, a total of 24 symptom (9 from YMRS, 12 from HAM-D, and 3 from SAPS) ratings were analyzed.

Statistical analysis

Categorical/nominal variables were summarized as frequency and percentage and were analyzed using Chi-square test. Continuous variables were expressed as mean and standard deviation and were analyzed using one-way analysis of variance test or Kruskal–Wallis test as appropriate. The level of significance was set at P < 0.05. All analysis was performed using statistical software IBM SPSS Statistics for Windows, Version 19.0. (trial version) Armonk, NY: IBM Corp.

RESULTS

A total of 162 patients were included in the final analysis. More than half (59.3%) of the patients were male. The three groups did not differ significantly regarding the gender distribution. Majority of the patients were from rural background (85.8%), married (74%), and belonged to lower socioeconomic status [Table 1].

Table 1

Sociodemographic characteristics of study population

Characteristics	EOBD (n=67), n (%)	IOBD (n=59), n (%)	LOBD (n=36), n (%)	Total (n=162), n (%)	χ2 (df)	P
Gender
Female	25 (37.3)	25 (42.4)	16 (44.4)	66 (40.7)	0.596 (2)	0.742
Male	42 (62.7)	34 (57.6)	20 (55.6)	96 (59.3)
Locality
Urban	5 (7.5)	17 (28.8)	1 (2.8)	23 (14.2)	16.69 (2)	<0.001
Rural	62 (92.5)	42 (71.2)	35 (97.2)	139 (85.8)
Marital status
Married	33 (49.2)	51 (86.4)	36 (100)	120 (74.1)	38.79 (2)	<0.001
Unmarried	34 (50.8)	8 (13.6)	0	42 (25.9)
Education
Up to primary	8 (11.9)	2 (3.4)	0	10 (6.1)	52.23 (6)	<0.001
Up to secondary	12 (17.9)	18 (30.5)	12 (33.3)	42 (25.9)
Above secondary	38 (56.7)	17 (28.8)	0	55 (34)
Illiterate	9 (13.4)	22 (37.3)	24 (66.7)	55 (34)
Occupation
Unemployed	33 (49.3)	25 (42.4)	14 (38.9)	72 (44.4)	21.27 (10)	0.019
Unskilled worker	6 (9)	8 (13.6)	6 (16.7)	20 (12.3)
Semi-skilled	3 (4.5)	2 (3.4)	8 (22.2)	13 (8)
Skilled worker	2 (3)	6 (10.2)	1 (2.8)	9 (5.6)
Clerical	21 (31.3)	18 (30.5)	7 (19.4)	46 (28.4)
Semi-profession	2 (3)	0	0	2 (1.2)
Profession	0	0	0	0
Monthly income (Rs.)
≤1589	4 (6)	2 (3.4)	0	6 (3.7)	25.040 (10)	0.005
1590-4726	15 (22.4)	12 (20.3)	16 (44.4)	43 (26.5)
4727-7877	28 (41.8)	25 (42.4)	8 (22.2)	61 (37.7)
7878-11816	12 (17.9)	12 (20.3)	12 (33.3)	36 (22.2)
11,817-15,753	0	4 (6.8)	0	4 (2.5)
15,754-31,506	8 (11.9)	4 (6.8)	0	12 (7.4)
≥31,507	0	0	0	0

P value calculated using Chi-square test. EOBD - Early-onset bipolar disorder; IOBD - Intermediate-onset bipolar disorder; LOBD - Late-onset bipolar disorder

There was a significant difference between the three subgroups regarding episodes per year; the early-onset group had significantly more episodes per year (P < 0.001) [Table 2]. The prevalence of family history of mood disorder was also significantly higher in the early-onset subgroup (P < 0.001) [Table 3].

Table 2

Comparison of the subgroups on age of presentation, onset, and episodes per years

	Mean (SD)			F (df)	P*
	EOBD	IOBD	LOBD
Age of presentation (years)	23.5 (8.3)	35.1 (10.3)	48.9 (5.4)	104.36 (161)	<0.001
Age at onset (years)	16.5 (2.6)	26.1 (3.2)	43.4 (5.4)	653.79 (161)	<0.001
Episodes per year	1.8 (1.9)	0.8 (0.8)	0.6 (0.5)	12.80 (161)	<0.001

*P value calculated by using one-way ANOVA test. SD - Standard deviation; EOBD - Early-onset bipolar disorder; IOBD - Intermediate-onset bipolar disorder; LOBD - Late-onset bipolar disorder; ANOVA - Analysis of variance

Table 3

Family history of mood disorder among the subgroups

Family history of mood disorder	EOBD, n (%)	IOBD, n (%)	LOBD, n (%)	Total, n (%)
Present	35 (52.2)	10 (16.9)	6 (16.6)	51 (31.5)
Absent	32 (47.8)	49 (83.1)	30 (83.4)	111 (68.5)

χ2=22.8 at df=2, P<0.001 (S). EOBD - Early-onset bipolar disorder; IOBD - Intermediate-onset bipolar disorder; LOBD - Late-onset bipolar disorder

Among these 35 affected families of early-onset probands, 51 first-degree relatives were found to be affected from mood disorder. Among these 51 first-degree relatives, 33 (64.7%) precipitated illness before the age of 21 years which indicates that they lay in the same early-onset group.

Manic, depressive, and psychotic symptoms revealed a significant difference in symptom profiles between the early, intermediate, and late-onset groups (P < 0.001). It was found that among different YMRS, HAM-D, and SAPS items, the EOBD group had significantly greater scores on irritability, motor activity–energy, sexual interest, depressed mood, delusions, and thought disorders than the other groups. Similarly, the LOBD group had a significantly greater score on elevated mood than the other groups [Table 4].

Table 4

Comparison of the three subgroups on different items of Young Mania Rating Scale, Hamilton Depression Rating Scale, and Scale for the Assessment of Positive Symptoms Scale (data presented as mean and standard deviation)

Scale	Items	EOBD (n=67)	IOBD (n=59)	LOBD (n=36)	F (df)	P
YMRS	Elevated mood	2.94 (0.92)	3.17 (0.87)	3.67 (0.68)	8.57 (161)	<0.001
	Increased motor activity-energy	3.14 (0.81)	2.23 (0.87)	1.78 (0.93)	34.11 (161)	<0.001
	Sexual interest	1.12 (0.95)	0.62 (0.66)	0.33 (0.48)	14.12 (161)	<0.001
	Sleep	3.07 (1.08)	3.22 (1.19)	3.33 (0.83)	0.74 (161)	0.477
	Irritability	3.88 (2.60)	2.31 (1.81)	1.16 (1.39)	21.20 (161)	<0.001
	Speech (rate and amount)	5.31 (1.83)	5.42 (2.04)	5.78 (1.49)	0.78 (161)	0.461
	Disruptive-aggressive behavior	3.49 (2.45)	2.92 (2.55)	3.33 (0.96)	1.04 (161)	0.357
	Appearance	0.69 (0.701)	1.25 (1.14)	0.93 (0.81)	5.99 (161)	0.003
	Insight	3.88 (0.41)	3.80 (0.48)	3.67 (0.68)	2.02 (161)	0.136
HAM-D	Depressed mood	0.93 (0.81)	0.64 (0.67)	0.39 (0.31)	7.89 (161)	<0.001
	Feeling of guilt	0.39 (0.60)	0.49 (0.65)	0.31 (0.53)	1.05 (161)	0.354
	Suicide	0.67 (1.09)	0.85 (1.23)	0.54 (1.04)	0.90 (161)	0.410
	Work and activities	0.94 (1.25)	0.83 (1.33)	1.11 (1.47)	0.50 (161)	0.610
	Retardation	0.37 (0.81)	0.37 (0.72)	0.46 (0.78)	0.19 (161)	0.827
	Anxiety psychic	0.60 (1.10)	0.75 (1.33)	0.49 (1.04)	0.58 (161)	0.559
	Anxiety somatic	0.34 (0.67)	0.51 (0.73)	0.37 (0.69)	1.00 (161)	0.368
	Somatic symptoms gastrointestinal	0.39 (0.72)	0.34 (0.61)	0.40 (0.74)	0.12 (161)	0.890
	Somatic symptoms general	0.55 (0.82)	0.32 (0.60)	0.60 (0.77)	2.18 (161)	0.117
	Genital symptoms	0.48 (0.68)	0.41 (0.65)	0.40 (0.65)	0.25 (161)	0.782
	Hypochondriasis	0.19 (0.47)	0.24 (0.47)	0.17 (0.45)	0.30 (161)	0.739
	Loss of weight	0.43 (0.49)	0.47 (0.50)	0.40 (0.49)		0.787
SAPS	Hallucinations	0.67 (1.34)	0.88 (1.52)	0.43 (1.01)	1.44 (161)	0.285
	Delusions	3.74 (2.93)	2.71 (2.41)	2.43 (2.32)	3.82 (161)	0.024
	Positive formal thought disorder	3.44 (3.12)	2.38 (2.11)	2.11 (1.56)	4.41 (161)	0.014

P value calculated using one-way ANOVA test. EOBD - Early-onset bipolar disorder; IOBD - Intermediate-onset bipolar disorder; LOBD - Late-onset bipolar disorder; YMRS - Young Mania Rating Scale; HAM-D - Hamilton Depression Rating Scale; SAPS - Scale for the Assessment of Positive Symptoms; ANOVA - Analysis of variance

DISCUSSION

BD patients were divided into three subgroups based on AAO (EOBD, IOBD, and LOBD), and a significant difference was found between the three subgroups on clinical presentation (different items of YMRS, HAM-D, and SAPS scale) and the presence of family history of the illness. Similarly, a previous study had divided BD patients into early-onset (age ≤20 years), intermediate-onset (age >20 to ≤30 years), and late-onset (age >30 years) subgroups. They found early-onset patients having more scores in hypersexuality, disinhibition, depressed mood, psychotic symptoms, and overall severity, whereas late-onset patients had greater scores on the Depression Rating Scale.[30] A study from Turkey also supported the hypothesis that adolescent-onset mania may be a different subtype than adult-onset mania with respect to type of episode, phenomenology, and clinical features.[31]

In the present study, depressed mood had a higher rating in the EOBD subgroup than the other groups, as was similarly reported by other studies from different countries.[3032] Another finding of the present study was that the psychotic symptoms, including delusions and thought disorder, were more common in the early-onset group, as was similarly reported by other researchers.[33] On the other hand, the LOBD group had a greater score on elevated mood. Supporting this, a study on geriatric population reported a greater average YMRS item rating for happiness and cheerfulness in elders with late onset compared to those with early-onset illness.[34]

In this study, early-onset patients experienced more number of episodes per year than the other groups, which indicates that they are more prone to be rapid cyclers. Supporting this, a previous study also reported that onset before 19 years of age eventually leads to the development of rapid cycling.[28]

Upon direct comparison of the EOBD and LOBD subgroups (the cutoff points were AAO before 18 years and after 40 years for the two subgroups), past literatures suggest that early-onset patients had more psychotic symptoms, mixed features, and family history of affective illness. EOBD and LOBD differ in clinical expression and familial risk and may therefore be considered to be different subtypes of manic-depressive illness.[33]

The family history of illness was significantly higher in the EOBD subgroup in the current study, and most of the first-degree relatives (64.7%) lied in the same proband (early onset) which supports the hypothesis that EOBD is a more heritable form of illness. Similarly, studies have found that bipolar siblings are much more likely to belong to the same theoretical AAO subgroup than to different subgroups.[1116] This finding confirms intrafamilial correlation for AAO in bipolar affective disorder. The existence of three “AAO subgroups” and the clustering of siblings in the same subgroup might be attributable to different genetic vulnerability factors supporting the view of genetic heterogeneity in BD.

CONCLUSION

AAO of BD is significantly associated with their clinical presentation and prevalence of family history of illness. The EOBD subgroup had significantly greater scores on several YMRS, HAM-D, and SAPS items, and greater prevalence of family history of illness can be considered as a specific phenotype of BD patients, which is more homogenous, severe, and heritable form of illness, could be a potential target for future genetic researches. AAO can be used a valid marker to identify homogenous subgroups in BD patients.

Limitations

Only patients of mania were included, and depressive patients were excluded for proper comparison of one aspect of illness. There is no standard definition of cutoff value of AAO, and the possibility of recall bias in reporting AAO could not be ruled out.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.